linear amp for HV (TC) / Electromagnetic Radiation / Forums

Forums

4hv.org :: Forums :: Electromagnetic Radiation

« Previous topic | Next topic »

linear amp for HV (TC)

1 2 3 next

Move Thread

LAN_403

aquilarubra

Fri Mar 19 2010, 10:49AM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

Hi,
I need help in building a linear amp. I was thinking perhaps to open a context with a prize if there is interest. Or anybody interested, just contact me.
I still try to understand what the best technology would be (valves, fast switching IGBTs, etc.). So, advises are appreciated.

Basically, I want a HV (1000-3000V) linear amp (class A), ideally about 50-400W in selectable stages of 50W. I should connect it to a generic signal generator. The output should feed a HV transformer, or a TC, or go directly into a plasma tube, which I use for music therapy research.

I need a frequency range of 20kHz-30MHz. I couldn't find anything off the shelf, since every linear amp on the market is designed only for hams, and starts at 1.8Mhz. There are some prohibitive industrial solution indeed, but nothing basic, as I need.

It should just use simple and easy to find components, and have the smallest number of components, without sacrificing stability.

So, please advise, or address me to any forum, company, etc. that could design such kind of amp. Thank you!

Steve Conner

Fri Mar 19 2010, 12:26PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

You're not going to get it, the frequency range you want is too wide to be easily doable.

Or at least, the industrial Class-A amps you saw are the cheapest way of doing it. If you asked me to design you an amp for this power and frequency range, I'd charge you enough that you'd be better off buying one of these amps from a used test equipment place. And I suspect any other designer good enough to do it would charge just as much.

Maybe you should build two amps, a 20kHz to 1.8MHz and a 1.8MHz to 30MHz.

Proud Mary

Fri Mar 19 2010, 01:49PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

It's not easy to see how an amplifier with such a wide bandwidth could be constructed.

Using valves, I would expect there to be at least three frequency ranges,
and probably four between 20kHz and 30MHz. The frequency range would be selected by means of a multi-pole
ceramic rotary wafer switch.

However you choose to do it, it will be expensive.

aquilarubra

Fri Mar 19 2010, 01:59PM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

I see. The industrial amps I saw went up to GHz... which is not what I need.

If it is easier, I need a minimum of 4MHz. So, it would be 20KHz-4MHz. I still need it though.

Thanks!

aquilarubra

Fri Mar 19 2010, 02:04PM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

Hey, also the 20kHz to 1.8MHz option seems a great idea!! But if it is not too complicate to push it up to 4MHz, it would be even better :)

Wouldn't IGBTs allow easier construction? I saw that it is easy to find 3000V IGBTs.

GeordieBoy

Fri Mar 19 2010, 03:01PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

Take a look at the product range from AR. These broadband power amplifiers are used for EMC compliance testing where it is desirable to perform one long sweep from 10kHz up to 30MHz or more.

They're ideal for experimental work because their Class A operation makes them almost impossible to blow up. They aren't cheap though.

-Richie,

radiotech

Fri Mar 19 2010, 04:32PM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

aquilarubra wrote:

"The output should feed a HV transformer, or a TC, or go directly into a plasma tube, which I use for music therapy research."

What wavelength does your tube produce? Are you using it for
light therapy?

aquilarubra

Fri Mar 19 2010, 05:13PM

Registered Member #2471 Joined: Wed Nov 18 2009, 07:23AM
Location:
Posts: 17

I was aware of those broadband amps... if I had that money, I would not be posting here :)

The wavelengths of the tubes are just what I put into the tube (so, 20kHz-30MHz). Usually sines. The light output has no importance, since I experiment with lower freqs. Usually they are not audible, but I still consider them as sound.

Mattski

Fri Mar 19 2010, 08:08PM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

I would think it may be feasible with a distributed amplifier, which is basically the same as a traveling wave tube amp. You can get extremely large bandwidths (theoretically) down to DC with such amplifiers. These amplifiers tend to be rather inefficient though, and across such a width bandwidth you would need to be rather careful about dispersion in your delay lines. You can't use transmission lines at this frequency so it will be pretty hard I think. Also, matching to your load will be quite challenging or impossible as its impedance will vary significantly with frequency.

If you're only doing a single frequency, or frequency ranges, at a time though there's absolutely no reason to use a single amplifier, it's much easier to use more highly optimized drivers across a mixture of frequency rangle.

radiotech

Sat Mar 20 2010, 06:32AM

Registered Member #2463 Joined: Wed Nov 11 2009, 03:49AM
Location:
Posts: 1546

Your specification of 3000 V 400 W would mean the output Z was 22,500. Now you want to increment the output in steps starting at 50 W. 50 W at 3000 V would give give you a Z of 180,000.
What exactly is the Z of the load at 50 W and 400 W: That would be the starting point. Decide if you want constant I, E or Z. The variable amplifier can only give you a choice of being a current or voltage source.

Perhaps if you described in detail the 'plasma tube'. I thought is was like a discharge lamp used with a spectrophotometer.

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob